A Probabilistic Matrix Factorization Method for Identifying lncRNA-disease Associations.

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• Long-non-coding RNAs (lncRNAs) are crucial in biological processes and serve as potential disease biomarkers.
• Traditional experimental validation of lncRNA-disease associations is time-consuming and costly.
• Computational models are increasingly used to predict these associations.

Purpose of the Study:

• To develop a novel computational model for predicting lncRNA-disease associations.
• To integrate lncRNA-miRNA, miRNA-disease, and lncRNA-disease networks for enhanced prediction.
• To validate the model's efficacy using established datasets and case studies.

Main Methods:

• Construction of lncRNA-miRNA, miRNA-disease, and lncRNA-disease association networks.
• Integration into a weighted lncRNA-disease association network, updated using KNN and semantic similarity.
• Development of the PMFILDA model based on probability matrix decomposition.
• Leave One Out Cross-Validation (LOOCV) for performance evaluation.

Main Results:

• The PMFILDA model demonstrated superior prediction performance compared to existing state-of-the-art methods.
• Case studies on breast cancer, lung cancer, and colorectal cancer confirmed PMFILDA's effectiveness.
• The integrated network approach and probability matrix decomposition yield accurate predictions.

Conclusions:

• PMFILDA offers a powerful and efficient computational approach for identifying lncRNA-disease associations.
• The model has significant potential for biomarker discovery in various diseases, including cancers.
• This study highlights the utility of integrated network analysis in bioinformatics.